LIGHTHOUSE, OR RADIO SIGNAL? Lynne at Knowledge Problem and Eric at Marginal Revolution are doing some heavy intellectual lifting on whether a satellite-driven, open-signal Global Positioning System qualifies as a pure public good. Neither post considers the possibility that such satellites, whether sending signals in the clear, or sending signals that require a special decoder -- and thus the possibility of excluding free riders by selling renewable code keys -- have elements of natural monopoly (or is it a transaction cost problem in which one satellite club can steal members from two or more smaller clubs??) that might vitiate contestability.
There is a precursor to the Global Positioning System that is kind of interesting. The Illustrated Longitude (the original book is a great read, there are marvelous maps, drawings, and photos in the illustrated version -- details or compare prices) tells of a plan to anchor ships at well-defined locations in the ocean that would fire guns to give the hour in London. Give or take the speed of sound, ships within earshot of the guideships would be able to reckon their longitude by comparing their observation of local noon with the hour gun signal they heard. (But if one could anchor guideships in the ocean, one could put identifying marks on them exactly as lighthouses and bell buoys, and reckoning longitude would cease to be a problem as ships could proceed from guideship to guideship.) Satellites can be anchored -- it's called geosynchronous orbit -- and run on batteries without getting lonely, thus addressing several practical problems that scuppered the guideship idea.